steve_v

So opting out of analytics doesn't actually opt out of analytics...Which was my complaint in the first place. I stand by my earlier "firewalling KSP was the correct choice".

Probably doable with MM. I'd start digging into the .cfg files for the stock powered wheels if I were you. Copy the relevant modules to a gear part and see what happens. That said, I don't really see the point of this; from a realism point of view (real aircraft use pushback tractors for this), or a from a gameplay perspective, as we already have powered wheels for powered rovers. Aircraft generally don't need to go backwards, and if retractable landing gear have no disadvantages over wheels people are very likely to use them on rovers rather than using the part meant for the job. It's a balance thing.

Today's complaint: Busybodies who write and distribute corporate feel-good "motivational" garbage, and passive-aggressive note-leavers in general. I'd love to post the latest puerile (attempt at poetry?) example, but it might just bite me in the ass... I'm sure you know what I'm talking about, it's everywhere. The fact that some chair-warmer is getting paid to tell people who do the core, physical work of the business, in the most condescending and childish manner possible, to be "Better Every Day", or "Improve Your Aim", makes me want to strangle them with their own gizzards... Or at the very least drag them out of their comfortable office and into the real world.

Right, so in order to opt out one has to submit a unique token to unity's servers, upon receipt of which (and subsequent clicks) they claim to stop collecting the data that KSP is obviously still sending... Do I need to post a packet trace to show that the game is still connecting to Unity servers despite opting out, or is logic alone enough to point out the obvious flaw in this system? Please provide an opt-out mechanism, or better yet an opt-in mechanism, that actually stops the application from sending data. I do mean any data, including my IP address, unique game identifier, and the fact that I have opted out. I firewalled this application long ago, because I don't trust SQUAD as far as I can throw them. If this is the best "opt-out" they can come up with, and that only after quite some outcry, it was clearly the correct decision. Why yes, yes it does. and it almost certainly transmits your game ID to see if it matches the opt-out list... So they still know who is playing, when, and where, at the very least. Hell, even simply pinging cloud.unity3d.com will tell them that. @SQUAD Please stop phoning home. We've asked, we've complained, you're still doing it. Stop. Even assuming the data you're sending is harmless, NO MEANS NO. I can't see exactly what you're sending, and I shouldn't have to firewall your products to be sure you're not going behind my back. You lost my trust in this matter some time ago by including redshell and removing the existing opt-out mechanism. The new one is a farce, and has done nothing whatsoever to regain it. After opting-out, KSP is still attempting to connect to, in no particular order: data-optout-service.uca.cloud.unity3d.com us-west-1.elb.amazonaws.com config.uca.cloud.unity3d.com cdp.cloud.unity3d.com I can tell you without doubt that packets are being exchanged with all of the above, but being tls1.2 I can't speculate as to their content. Getting a connection to data-optout-service.uca.cloud.unity3d.com, and presumably an "opt-out" response, does not prevent it from connecting to the rest. As the encrypted data transferred to cdp.cloud.unity3d.com is approximately the same size whether the opt-out service is reachable or not, I can only assume that opting out does not prevent data being transmitted. Pretty liquid-poor effort if you ask me. If you really want the whole trace, I assume you have a convenient mechanism to share for the decryption of these tls packets? Do you have concrete proof that KSP is not phoning home once you opt out?

Don't suppose anybody snagged this file before it vanished? Looks like the dropbox link expired some time ago.

Ahh, the male non-maskable interrupt. A guaranteed reboot if triggered with sufficient force. I wouldn't call it funny, but today I got screwed over by the bass-ackwards 'murican date format. Again.

Because they decided to check other services with your username, or (more likely) because a bot got around to it today and it's going through it's list of sites. Expect this to happen every now and then, it's the nature of the internet. My home server weathers (and auto-firewalls) dozens of bot-generated cracking attempts per day, and it's the epitome of obscure. No big deal if your passwords (and password recovery systems) and/or 2FA are strong. And do not reuse passwords, especially with the same username. <complaint> IMO "password recovery" systems are a security risk in general, particularly as a great many sites still use "secret answers" that are easily scraped from publicly accessible information. Just remember your damn (complex and unique) passwords already... Or use a decent password manager. Account recovery via phone call and stupid questions (which anyone who has access to your garbage can answer anyway) is even worse, as it is ripe for social engineering attacks. Many banks still allow this, mine included. </complaint>

Bring it. A spade is a spade, and this one is (still) broken. The only fanboi attacks I've weathered so far are variations of "Linux users are a minority, so Linux bugs won't/shouldn't get fixed", an argument I am not about to buy. Oh, I forgot: "It's probably hard, cut them some slack" (written by people with no programming experience of course), I get that one too.

Regression introduced in 1.4.0. 5 month old bug report, no apparent progress despite 5 patches. Primary reason I'm neither playing KSP at the moment nor buying the DLC. Short answer: Yes. Completely FUBAR. Squad and/or Unity completely borked USB HID input on GNU/Linux when switching to SDL2, and the touted "steam controller" support isn't usable either. Make of it what you will, but your joystick ain't going to work properly (or at all in most cases) with any 1.4.x release. Just in case someone at @SQUAD reads this: Thanks, thanks a bunch for this horribly broken update. I'll be sticking with 1.3.x (and no DLC) until you fix the mess you made. As demonstrated here, 1.4.5 still isn't it. Yeah, AFBW works just fine... and makes squad look somewhat incompetent in the process, by fixing their bugs and generally leaving the stock input system for dead. What it does should have been implemented in the game engine a long time ago.

If you have read this thread, and understood at least 80% of what each and every post was talking about, you've probably already realised that it's a whole lot easier and more productive to learn an existing language than it is to write a new one. I can see the appeal as a learning exercise, but for practical use, why?

I wonder if there is a rate-limit on that... Tempting to script up a wee test.

Excellent. I pick up my mouse with thumb and finger thusly when dealing with limited space. So long as it doesn't require a Windoze-only utility to do so... Guarantee it will, and they'll probably be pointless blindingly-bright RGB gimmicks. I'll take zero LEDs thanks, besides those required for the optics. If I can get at the PCB I will gladly cut them off, but I still faintly resent paying for the things in the first place. Optical switches sound like a nice idea, and should be relatively inexpensive to build. Traditional microswitches are the first thing to fail IME, and the first evidence of this, contact bouncing, is incredibly annoying. The bit touting optical scroll-wheel sensors as some kind of revolution is a bit odd though, a great many el-cheapo mice already have an optical gate as a scroll sensor. I sure hope the thing is well sealed, cleaning lint and dust (and probably ash, cough cough) out of optical switches isn't going to be any more entertaining than the same for their mechanical counterparts. If the overall quality is good (i.e all those finicky, clippy plastic bits stay put), I might give it a go... despite it's still falling into the "gimmicky" problem that seems to plague gaming mice. I don't want a lego mouse, but if that's what it takes to get away from this hip "ergonomic" right-handed thing, so be it. Sure as hell not backing it pre-production though, at least not at that price. It might, and I'll probably try it out if the price of the final product is reasonable. As for "white whales", I have enough (black as it happens) cetaceans to last me a few years. Worst-case scenario, I'll have to replace some microswitches and skates to keep them alive.

Probably not difficult to do with ModuleManager, since it's just a case of messing with part configs. I don't see the point though. If you want more thrust, just use a bigger engine. All this would achieve is changing extra engine mass for more thrust to extra radiator mass for more thrust. It's also unrealistic and silly, IMO. There's already a better solution wrt nukes: The LANTR in Atomic Age. More thrust at the expense of efficiency and oxidiser.

I'd back a KSP2 on one condition: No Unity3D. I'd settle for a KSP 1.x free of unity-related recurring bugs and general jankiness.

Why yes, yes they are. Not a particularly serious problem IMO, but certainly not an improvement.

This forum, closely followed by Spacedock and KerbalX. Sometimes github or ksp.sarbian.com too.

Really wish people would stop spouting this particular "advice". "Run it as admin" is almost never a solution, and almost always terrible security practice. If permissions are screwey, fix them. No, it's not. A picture of a generic crash dialog is totally useless, please upload your logs and post a link.

I acquired a stash of new-old stock 1.1a intellimice(?) from ebay a while back, after failing miserably to find a worthy replacement for the one I wore down the buttons on. I've tried at least 5 different "gaming" mice, and frankly, they all suck. Gimmicky, fragile, heavy, and the wrong shape. I particularly hate "handed" mice, as I tend to use whichever hand is nearer at the time. Might try one of these new ones, if they are indeed the same shape and layout. IMO the original is the best mouse ever made, and the only worthwhile product Microsoft ever made. The only thing I didn't like about them was the unnecessary rear-facing LED in the base, but those are easily removed. I see the new one has an even more superfluous and annoying white LED, which I will also have to eviscerate. Ed. Wait... Gah! The new intellimouse is not symmetrical, it's "right handed". Screw it then, not a replacement for the 1.1/1.1a at all. Honestly, why can nobody do a simple, reliable, symmetrical mouse any more? This: Is how you do it.

It could also be anything that interferes with the rendering path, such as GPU monitoring overlays or shader injectors that hook or override the DX dlls. @Defectum: You don't happen to have RivaTuner or some other GPU utility running, do you? They have been known to cause KSP to crash at startup.

Get a powered hub and you will quickly discover whether or not it is a power supply problem. A USB3 PCIE card with it's own internal power connector would do the same. Measuring the current delivered (possibly into a dummy load) would also be worthwhile, if the USB ports fail to meet spec you have a bulletproof warranty claim. Devices to do this are cheap, or you could just butcher a USB cable and use a multimeter. USB ports should not get hot when delivering rated current. Devices plugged into them might though - if you haven't already, get that power hungry wifi dongle away from the port on an extension cable and see where the heat is actually coming from. Check out your cooling in general too, it's just possible there's enough heat coming from something nearby on the motherboard to kill the USB controller or voltage regulator when under load. I've seen laptops that have the GPU heatpipes practically on top of the USB ports. What kind of machine is this anyway? USB2 spec is only 500mA. USB3: 900, unless they implement one of the charging port standards. Most are indeed good for ~2A, but it would pay to check the hardware specifications - particularly if talking about a laptop or rear ports on a cheap motherboard.

I don't see the benefit of a minimum character requirement, but I would like to see a requirement for help requests in the tech subs to be written in comprehensible sentences with punctuation and linebreaks.

Uhh, whut? Doesn't look "lost" to me: As for "rocket crashed into page 1"... There is no official 1.4 release, on the forum or on github. You might try compiling it (possibly pre the revert commit) against 1.4 or using one of the forks mentioned in the forum thread. If you want to know why ferram4 reverted those commits, you'll have to ask him.

Don't be so sure...

I don't recognise the part name, but you could try uploading the craft to KerbalX, to see what its mod detection makes of it.

Count me as exactly this ^.